Only include dyn Trait<Assoc = ...> associated type bounds for Self: Sized associated types if they are provided

[compiler-errors]

Since #136458, we began filtering out associated types with Self: Sized bounds when constructing the list of associated type bounds to put into our dyn Trait types. For example, given:

trait Trait {
    type Assoc where Self: Sized;
}

After #136458, even if a user writes dyn Trait<Assoc = ()>, the lowered ty would have an empty projection list, and thus be equivalent to dyn Trait. However, this has the side effect of no longer constraining any types in the RHS of Assoc = ..., not implying any WF implied bounds, and not requiring that they hold when unsizing.

After this PR, we include these bounds, but (still) do not require that they are provided. If the are not provided, they are skipped from the projections list.

This results in dyn Trait types that have differing numbers of projection bounds. This will lead to re-introducing type mismatches e.g. between dyn Trait and dyn Trait<Assoc = ()>. However, this is expected and doesn't suffer from any of the deduplication unsoundness from before #136458.

We may want to begin to ignore thse bounds in the future by bumping unused_associated_type_bounds to an FCW. I don't want to tangle that up into the fix that was originally intended in #136458, so I'm doing a "fix-forward" in this PR and deferring thinking about this for the future.

Fixes #140645

r? lcnr

[rustbot]

HIR ty lowering was modified

cc @fmease

[lcnr]

it's not rly needed_associated_types anymore, is it. It's ordered_associated_types or whatever. Pls rename the variable and add a comment explaining the approach of

• collect all associated types in a stable order by walking over the dyn trait with identity substitutions
• then use that to order user annotations
• then with elaborated super bounds, not replying user annotations